Chapter XIV: The Geographical Development of Coast-Lines 393 (12)

Yes, but if we conceive the submergence of the Isthmus of Panama to coincide with great elevation of northern lands, would not such geographical conditions bring about a glacial epoch comparable to that of Pleistocene times? It is hard to see how they could. No doubt the climate of all those regions that would be affected by the withdrawal of the Gulf Stream alone would become still more deteriorated if they stood some 3000 feet higher than now. A vast area in the north-west of Europe would certainly be uninhabitable, but it is for the advocates of the "earth-movement hypothesis" to explain why those inhospitable regions should necessarily be covered with an ice-sheet. For the production of great snow-fields and continental ice-sheets, considerable precipitation, no less than a low temperature, is requisite. Under the conditions we have been imagining, however, precipitation would probably be much less than it is at present. But to whatever extent north-west Europe might be glaciated, it is obvious that the geographical revolutions referred to could have little influence on the climate of south-eastern Europe, not to mention central and eastern Asia. Nor could they possibly influence the climate of the Pacific coast-lands of North America. And yet, as is well known, the climate of all those regions was more or less profoundly affected during the Glacial period. To account for the widespread evidences of glaciation by means of elevation it would therefore seem necessary to infer that all the affected areas were in Pleistocene times uplifted _en masse_ into the Arctic zone that stretches above our heads. Now it seems easier to believe that the snow-line was lowered by several thousand feet than that the continents were elevated to the same extent. Glaciation, as we have seen, was developed in the same directions and over the same areas as we should expect it to be were the snow-line to be generally depressed. To put it in another way, were the snow-line by some means or other to be lowered over Europe, Asia, and North America, then, with sufficient precipitation, great ice-fields and glaciers would reappear in the very regions which they visited during Pleistocene times. Neither elevation nor depression of the land would be required to bring about such a result. Certain advocates of the "earth-movement hypothesis," however, do not maintain that all the glaciated areas were uplifted at one and the same time. The glaciation of the Alps, they think, may have taken place earlier or later than that of north-western Europe, while the ice-period of the Rocky Mountains may not have coincided with that of eastern North America. It is not impossible, they suppose, that the glaciation of the Himalayas may have been caused by an uplifting of that great chain, quite independent of similar earth-movements in other places. It can be demonstrated, however, that the glaciation of the Alps and of northern Europe were contemporaneous, and the facts go far to prove that the glaciers of the Rocky Mountains and the inland-ice of north-east America likewise co-existed. At all events all the old glacial accumulations of our hemisphere are of Pleistocene age, and it is for the advocates of the hypothesis under review to prove that they are not contemporaneous. Their doubts on the subject probably arise from the simple fact that they are well aware how highly improbable or even impossible it is that all those glaciated lands could have been pushed up within the snow-line at one and the same time.

Let me, however, advance to another objection. We know that the Glacial period was interrupted by at least one interglacial epoch of temperate and even genial conditions. Two glacial epochs with one protracted interglacial epoch are now generally admitted. How do the supporters of the "earth-movement hypothesis" explain this remarkable succession of climatic changes? Their views as to the cause of glacial conditions we have considered. If we can believe that the glacial phenomena were due to elevation of the land, then we need have no difficulty in understanding how glacial conditions would disappear when the continents again subsided to a lower level. Not only did North America and Europe lose all their early glacial elevation, but by a lucky coincidence the Isthmus of Panama reappeared, and the Gulf Stream resumed its beneficent course into the North Atlantic. This we are to suppose was the cause of the interglacial epoch. But I would point out that the geographical conditions which are thus inferred to have brought about the disappearance of the glacial climate, and to have ushered in the interglacial epoch, are precisely those that now obtain--and, nevertheless, we are not yet in the enjoyment of a climate like that of interglacial times. The strangely equable conditions that permitted the development of the remarkable Pleistocene flora and fauna are not experienced in the Europe of our day. And what about the second glacial epoch? Are we to suppose that once more the lands were greatly uplifted, and that convenient Isthmus of Panama was again depressed? Did the Alps, the Pyrenees, and the plateau of central France--in all of which we have distinct evidence of at least two glacial epochs--did these heights, one may ask, rise up to bring about their earlier glaciation, sink down again to induce interglacial conditions, and once more become uplifted at the succeeding cold epoch, to subside eventually in order to cause a final retreat of their glaciers?

But the climatic changes to be accounted for were in all probability more numerous and complex than those just referred to. Competent observers have adduced unmistakable evidence of three epochs of glaciation in the Alpine Lands of Europe. And we are not without distinct hints that similar changes have taken place in northern and western Europe. Nor in this connection can we ignore the evidence of several interglacial episodes which Mr. Chamberlin and others have detected in the glaciated tracts of North America. Even this is not all, for the upholders of the "earth-movement hypothesis" have still further to account for the climatic oscillations of post-glacial times. If it be hard enough to allow the possibility of one great movement of elevation having affected so enormous an area of our hemisphere, if we find it extremely difficult to believe either that one such widespread movement, or that a multitude of local movements, each more or less independent of the other, could have lifted the glaciated regions successively within reach of the snow-line--we shall yet find it impossible to admit that such remarkable upheavals could be repeated again and again.

We seem driven to conclude, therefore, that the "earth-movement hypothesis" fails to explain the phenomena of Pleistocene times. One cannot deny, indeed, that glaciation might be induced locally by elevation of the land. It is quite conceivable that mountains now below the limits of perennial snow might come to be ridged up to such an extent as to be capable of sustaining snow-fields and glaciers. And such local movements may possibly have happened here and there during the long-continued Pleistocene period. But the glacial phenomena of that period are on much too grand a scale and far too widely distributed to be accounted for in that way. And if the occurrence of even one glacial epoch cannot be thus explained, we may leave the supporters of the "earth-movement hypothesis" to show us what light is thrown by their urim and thummim on the origin of succeeding interglacial and glacial climates.

There is yet another physical condition of the Pleistocene and post-glacial periods which any adequate explanation must embrace. I refer to the oscillation of sea-level, of which so many proofs are forthcoming. It is very remarkable that almost everywhere throughout the maritime regions of formerly glaciated areas we find evidence of submergence. So commonly is this the case, that geologists have long suspected that the connection between glaciation and submergence might be one of cause and effect. The possible influence of great ice-sheets in disturbing the relative level of land and sea is a question, therefore, of very great importance. It is one, however, which must be solved by physicists. Croll and others have advocated the view that the great accumulations of ice of the Glacial period may have displaced the earth's centre of gravity, and thus caused the sea to rise upon the glaciated hemisphere. The various results arrived at by physicists are hardly comparable, because each has used different data, but it seems probable that we have in this view a _vera causa_ of oscillations of the sea-level. Another hypothesis would explain the rise of the sea as due to the attractive influence of the great ice-masses, but Dr. Drygalski's and Mr. Woodward's elaborate investigations would seem to have demonstrated that this notion does not account for the facts. Yet another speculation has been advanced. Mr. Jamieson has suggested that the mere weight of the ice-sheets would suffice to press down the earth's crust into a supposed liquid substratum, and this explanation has met with much acceptance. Unfortunately our knowledge of the condition of the earth's interior is so very limited that we cannot be certain as to how the crust would be affected by the weight of an ice-sheet. No doubt Mr. Jamieson's hypothesis gives a specious explanation of certain geological phenomena, but if there be no liquid substratum underlying a thin crust it cannot be true. At present the prevalent view of physicists appears to be that the earth is substantially solid. Professor George Darwin has shown that the prominent inequalities of the earth's surface could not be sustained unless the crust be as rigid as granite for a depth of 1000 miles. "If the earth be solid throughout," he remarks, "then at 1000 miles from the surface the material must be as strong as granite. If it be fluid or gaseous inside, and the crust 1000 miles thick, that crust must be stronger than granite, and if only 200 or 300 miles in thickness, much stronger than granite." This conclusion is obviously strongly confirmatory of Sir William Thomson's view, that the earth is solid throughout. But many geologists find it hard to account for the convolutions of strata and other structural phenomena on the supposition that the earth is entirely solid, and they are inclined, therefore, to adopt the hypothesis of a sub-crust layer of liquid matter. Whether this be actually the condition or not physicists must be left to determine. All that we need note is, that if there be any force in Professor Darwin's argument, it is obvious that the crust is possessed of great rigidity, and could not be readily deformed by the mere weight of an ice-sheet. According to Dr. Drygalski, however, the presence of an ice-sheet, by reducing the temperature of the underlying crust, would bring about contraction, and in this way cause the surface to sink. When the ice-sheet had disappeared, then free radiation of earth-heat would be resumed, the depressed isogeotherms would rise, and a general warming of the upper portion of the lithosphere would take place. But the space occupied by the depressed section, owing to the spheroidal form of the earth, would be smaller than that which it occupied before sinking had commenced, and consequently when the ice vanished expansion of the crust would follow, and the land-surface would then rise again. The whole question is one for physicists to decide upon, but I may point out that if Drygalski's explanation be well founded, then it is obvious that it throws no light upon the origin and subsequent disappearance of an ice-sheet. Somehow or other this ice-sheet comes into existence, and the cooling and contracting crust sinks below it; and that depressed condition of the glaciated area must continue so long as the ice-sheet remains unmelted. Re-elevation can only take place when, owing to some other cause or causes, the climate changes and the ice-sheet vanishes.

Those who advocate the "earth-movement hypothesis" as an explanation of the origin of extensive glaciation have welcomed Mr. Jamieson's view as harmonising well with their conclusions. They contend, as we have seen, that glacial conditions were induced by an extensive upheaval of the crust in northern latitudes, accompanied by a depression of the Isthmus of Panama. They then proceed to point out that the ice-sheets brought about their own dissolution by pressing down the crust, and introducing with submergence a disappearance of glacial conditions. See now how much they take for granted. In the first place, they assume an amount of pre-glacial or early glacial elevation of northern regions for which not a scrap of evidence can be adduced, while they can give no proof of contemporaneous depression of the Isthmus of Panama. Next, relying on Mr. Jamieson's hypothesis, they take for granted that the ice-sheets, called into existence by their postulated earth-movements, succeeded in depressing the earth's surface even below its present level. That is to say, the land, which, according to them, was in glacial times some 3000 feet higher than now, sank down under the weight of its glacial covering for, say, 3600 feet in north-western Europe. In North America, in like manner, all the pre-glacial elevation was lost--the land sinking below its present level for some 200 feet in New England, for 520 feet at Montreal, for 1000 to 1500 feet in Labrador, and for 1000 to 2000 feet in the Arctic regions. Now, even if we concede the reasonableness of Mr. Jamieson's hypothesis, and admit that a certain degree of deformation may take place under the mere weight of an ice-sheet, it is difficult to believe that the crust can be so readily deformed as the supporters of the "earth-movement hypothesis" seem to imply. If it could yield so readily to pressure, one is at a loss to understand how a great ice-sheet could accumulate--the ice would simply float off as the land subsided. Take the case of north-western Europe. The ice-sheet that covered Scotland did not attain, on the average, 3000 feet in thickness, and yet we are to suppose that it was able to depress the land for some 600 feet below its present level--that is to say, for 3600 feet below its assumed pre-glacial elevation. Either the ice depressed the crust to that remarkable extent, or the land upon which the ice accumulated was not nearly so high as the advocates of the "earth-movement hypothesis" have supposed. But the average I have taken for the thickness of the Scottish ice-sheet is excessive, for it was only in the low-grounds that the _mer de glace_ attained such a depth. A large part of our country, however, is mountainous, and the mountain-tops were, of course, not nearly so thickly mantled with ice as the valleys. And the same to even a larger extent holds good for the Scandinavian peninsula. If we take the thickness of the Scandinavian ice-sheet that coalesced with that of Scotland as 4000 feet, we shall be over the mark. Now, I ask, is it possible to believe that a sheet of ice of that thickness actually pressed down the crust of the earth for not less than 3600 feet? But if we accept the "earth-movement hypothesis," as it has been recently advocated, that is what we must believe. If we cannot do so, then we cannot accept the assumption of great elevation of the land in pre-glacial and glacial times. Let me put the case shortly: if the glacial marine beds and raised beaches of the Atlantic borders of Europe and North America owe their origin to depression induced by the weight of an ice-sheet, then it is quite certain that at the advent of glacial conditions the land could not have been so highly elevated as the advocates of the "earth-movement hypothesis" suppose. But if we are to accept the notion of great elevation of the land, then we must conclude that the submergence to which the raised beaches testify cannot have been caused by the pressure of ice-sheets.

It is hardly necessary to pursue this particular subject further, but before leaving it, attention may be drawn for a moment to the curious conclusion that the ice-sheets were self-destructive. One is left to guess at what particular stage the sinking process began, but if the earth's crust were as readily deformed as the extreme views I have been examining would compel one to imply, then depression must have commenced almost immediately with the accumulation of snow and ice. The several ice-sheets must soon have attained their maximum thickness, and their disappearance must have been correspondingly rapid. And yet all the evidence goes to show that a glacial epoch endured for a comparatively long time--for a time sufficient to account for a prodigious amount of rock-erosion, and for the accumulation of vast sheets of glacial débris and fluvio-glacial detritus.[BY]

[BY] It must not be inferred from the above remarks that I deny the possibility of deformation of the crust having been induced by the old ice-sheets. The geological evidence is certainly suggestive of such having been the case. But I much doubt whether the sinking of the surface was brought about by the mere weight of the ice pressing the crust down into a subjacent liquid layer. Dr. Drygalski's explanation would better account for the geological phenomena, but, according to Rev. Osmond Fisher, it cannot be maintained.

If it be difficult to understand how the "earth-movement hypothesis" can account for the origin of one glacial epoch, the difficulty is not lessened when we remember that there are two or more such epochs to account for. And until the advocates of that hypothesis can furnish us with some reliable evidence, they can hardly expect us to believe in their mysterious upheavals and depressions of northern and temperate regions, and in the no less wonderfully rhythmic movements of the Isthmus of Panama. In fine, the views which I have been controverting seem to me to be untenable, inasmuch as they are founded on mere assumptions, and do not even give a reasonable and intelligible explanation of the phenomena of glaciated regions, while they practically ignore or leave unsolved the problem of interglacial conditions.

Some five-and-twenty years have now elapsed since my lamented friend and colleague, James Croll, published his well-known physical theory of the Glacial period. That theory, as you all know, has been frequently criticised by physicists and others, to whose objections Croll made a final reply in his _Climate and Cosmology_. In that work he has successfully defended his views, and even added considerably to the strength of his general argument. I am not aware that since then any serious objections to Croll's theory have appeared. The only one indeed that seems to have attracted attention is that which has been urged especially by certain American geologists. Their belief is that the close of the Glacial period must have taken place at a much more recent date than Croll has inferred. And this belief of theirs is based upon various estimates which have been made as to the time required for the erosion of valleys and the accumulation of alluvial deposits since the Glacial period. Thus, according to Mr. Gilbert, the post-glacial gorge of Niagara, at the present rate of erosion, must have been excavated within 7000 years; while Mr. Winchell, from similar measurements of the post-glacial erosion of the Falls of St. Anthony, concludes that 8000 years have elapsed since the close of the Ice Age. I might cite a number of similar estimates that tend to show that since the close of the Glacial period only 7000 or 10,000 years have elapsed. What will archæologists say to this conclusion? We know that Egypt was already occupied by a civilised people nearly 6000 years ago, and their marvellously advanced civilisation at that time presupposes, according to Egyptologists, many thousands of years of development. Are we, then, prepared to admit that the close of the Ice Age coincided with the dawn of Egyptian civilisation? But all American observers are not so parsimonious with regard to post-glacial time. Thus Professor Spencer has given the age of the Falls of Niagara as 24,000 years, and he informed me recently that this does not represent half of the time since the formation of the third great series of glacial deposits of the Canadian uplands. In our own Continent similar estimates have been based on the rate of erosion of river-valleys, the rate of accumulation of alluvial deposits, of peat-bogs, of stalagmite in caves, and what not, with results that, to say the least, are rather discordant. The fact is that all such measurements and estimates, however carefully conducted and cautiously made, are in the nature of things unreliable. We are insufficiently acquainted with all the factors of the problem to be solved, and I cannot therefore agree with those who attribute much weight to conclusions based on such uncertain data. Dr. Croll's theory may eventually be modified, but I feel sure that it will not be overturned by the inconclusive and unsatisfactory estimates to which I have referred. Moreover, opponents of that theory may be reminded that its truth does not rest on the accuracy of its author's conclusion as to the date of the last Ice Age. That periods of high eccentricity of the earth's orbit have occurred is beyond all doubt, but whether the formulæ employed by Croll in calculating the date of the last great cycle can be relied upon for that purpose is quite another question. At present, so far as I understand the facts, the glacial and the interglacial phenomena are explained by the astronomical theory, and by no other. It gives a simple, coherent, and consistent interpretation of the climatic vicissitudes of the Pleistocene and post-glacial periods, and in especial it is the only theory that throws any light on the very remarkable climates of interglacial times.

X.

The Glacial Succession in Europe.[BZ]

[BZ] _Trans. Royal Soc. Edinburgh_, vol. xxxvii. (1892).

For many years geologists have recognised the occurrence of at least two boulder-clays in the British Islands and the corresponding latitudes of the Continent. It is no longer doubted that these are the products of two separate and distinct glacial epochs. This has been demonstrated by the appearance of intercalated deposits of terrestrial, freshwater, or, as the case may be, marine origin. Such interglacial accumulations have been met with again and again in Britain, and they have likewise been detected at many places on the Continent, between the border of the North Sea and the heart of Russia. Their organic contents indicate in some cases cold climatic conditions; in others, they imply a climate not less temperate or even more genial than that which now obtains in the regions where they occur. Nor are such interglacial beds confined to northern and north-western Europe. In the Alpine Lands of the central and southern regions of our Continent they are equally well developed. Impressed by the growing strength of the evidence, it is no wonder that geologists, after a season of doubt, should at last agree in the conclusion that the glacial conditions of the Pleistocene period were interrupted by at least one protracted interglacial epoch. Not a few observers go further, and maintain that the evidence indicates more than this. They hold that three or even more glacial epochs supervened in Pleistocene times. This is the conclusion I reached many years ago, and I now purpose reviewing the evidence which has accumulated since then, in order to show how far it goes to support that conclusion.

In our islands we have, as already remarked, two boulder-clays, of which the lower or older has the wider extension southwards, for it has been traced as far as the valley of the Thames. The upper boulder-clay, on the other hand, does not extend south of the midlands of England. In the north of England, and throughout Scotland and the major portion of Ireland, it is this upper boulder-clay which usually shows at the surface. The two clays, however, frequently occur together, and are exposed again and again in deep artificial and natural sections, as in pits, railway-cuttings, quarries, river-banks, and sea-cliffs. Sometimes the upper clay rests directly upon the lower; at other times they are separated by alluvial and peaty accumulations or by marine deposits. The wider distribution of the lower till, the direction of transport of its included erratics, and the trend of the underlying _roches moutonnées_ and rock-striæ, clearly show that the earlier _mer de glace_ covered a wider area than its successor, and was confluent on the floor of the North Sea with the Scandinavian ice-sheet. It was during the formation of the lower till, in short, that glaciation in these islands attained its maximum development.

The interglacial beds, which in many places separate the lower from the upper till, show that after the retreat of the earlier _mer de glace_ the climate became progressively more temperate, until eventually the country was clothed with a flora essentially the same as the present. Wild oxen, the great Irish deer, and the horse, elephant, rhinoceros, and other mammals then lived in Britain. From the presence of such a flora and fauna we may reasonably infer that the climate during the climax of interglacial times was as genial as now. The occurrence of marine deposits associated with some of the interglacial peaty beds shows that eventually submergence ensued; and as the shells in some of the marine beds are boreal and arctic forms, they prove that cold climatic conditions accompanied the depression of the land. To what extent the land sank under water we cannot tell. It may have been 500 feet or not so much, for the evidence is somewhat unsatisfactory.

The upper boulder-clay of our islands is the product of another _mer de glace_, which in Scotland would seem to have been hardly less thick and extensive than its predecessor. Like the latter, it covered the whole country, overflowed the Outer Hebrides, and became confluent with the Scandinavian inland-ice on the bed of the North Sea. But it did not flow so far to the south as the earlier ice-sheet.

It is well known that this later _mer de glace_ was succeeded in our mountain-regions by a series of large local glaciers, which geologists generally believe were its direct descendants. It is supposed, in short, that the inland-ice, after retreating from the low-grounds, persisted for a time in the form of local glaciers in mountain-valleys. This view I also formerly held, although there were certain appearances which seemed to indicate that, after the ice-sheet had melted away from the Lowlands and shrunk far into the mountains, a general advance of great valley-glaciers had taken place. I had observed, for example, that the upper boulder-clay is often well developed in the lower reaches of our mountain-valleys--that, in fact, it may be met with more or less abundantly up to the point at which large terminal moraines are encountered. More than this, I had noticed that upland valleys, in which no local or terminal moraines occur, are usually clothed and paved with boulder-clay throughout. Again, the aspect of valleys which have been occupied by large local glaciers is very suggestive. Above the point at which terminal moraines occur only meagre patches of till are met with on the bottoms of the valleys. The adjacent hill-slopes up to a certain line may show bare rock, sprinkled perchance with erratics and superficial morainic detritus; but above this line, if the acclivity be not too great, boulder-clay often comes on again. These appearances are most conspicuously displayed in the southern Uplands of Scotland, particularly in south Ayrshire and Galloway, and long ago they led me to suspect that the local glaciers into which our latest _mer de glace_ was resolved, after retreating continuously towards the heads of their valleys, so as to leave the boulder-clay in a comparatively unmodified condition, had again advanced and ploughed this out, down to the point at which they dropped their terminal moraines. Subsequent observations in the Highlands and the Inner and Outer Hebrides confirmed me in my suspicion, for in all those regions we meet with phenomena of precisely the same kind. My friends and colleagues, Messrs. Peach and Horne, had independently come to a similar conclusion; and the more recent work of the Geological Survey in the north-west Highlands, as they inform me, has demonstrated that after the dissolution of the general ice-sheet underneath which the upper boulder-clay was accumulated, a strong recrudescence of glacial conditions supervened, and a general advance of great valley-glaciers took place--the glaciers in many places coalescing upon the low-grounds to form united _mers de glace_ of considerable extent.

The development of these large glaciers, therefore, forms a distinct stage in the history of the Glacial period. They were of sufficient extent to occupy all the fiords of the northern and western Highlands, at the mouths of which they calved their icebergs, and they descended the valleys on the eastern slopes of the land into the region of the great lakes, at the lower ends of which we encounter their outermost terminal moraines. The Shetland and Orkney Islands and the Inner and Outer Hebrides at the same time nourished local glaciers, not a few of which flowed into the sea. Such, for example, was the case in Skye, Harris, South Uist, and Arran. The broad Uplands of the south were likewise clothed with snow-fields that fed numerous glaciers. These were especially conspicuous in the wilds of Galloway, but they appeared likewise in the Peeblesshire hills; and even in less elevated tracts they have left more or less well-marked traces of their former presence.

It is to this third epoch of glaciation that I would assign the final scooping out of our lake-basins and the completion of the deep depressions in the beds of our Highland fiords. All the evidence, indeed, leads to the conviction that the epoch was one of long duration.

It goes without saying that what holds good for Scotland must, within certain limits, hold good also for Ireland and England. In Wales and the Cumberland lake district, and in the mountain-regions of the sister island, we meet with evidence of similar conditions. Each of those areas has obviously experienced intense local glaciation subsequent to the disappearance of the last big ice-sheet.

Attention must now be directed to another series of facts which help us to realise the general conditions that obtained during the epoch of local glaciation. In the basin of the estuary of the Clyde, and at various other places both on the west and east coasts of Scotland, occur certain clays and sands, which overlie the upper boulder-clay, and in some places are found wrapping round the kames and osar of the last great ice-sheet. These beds are charged with the relics of a boreal and arctic fauna, and indicate a submergence of rather more than 100 feet. In the lower reaches of the rivers Clyde, Forth, and Tay the clays and sands form a well-marked terrace, and a raised sea-beach, containing similar organisms, occurs here and there on the sea-coast, as between Dundee and Arbroath, on the southern shores of the Moray Firth, and elsewhere. When the terraces are traced inland they are found to pass into high-level fluviatile gravels, which may be followed into the mountain-valleys, until eventually they shade off into fluvio-glacial detritus associated with the terminal moraines of the great local glaciers. It is obvious, in short, that the epoch of local ice-sheets and large valley-glaciers was one also of partial submergence. This is further shown by the fact that in some places the glaciers that reached the sea threw down their moraines on the 100-feet beach. It must have been an epoch of much floating ice, as the marine deposits contain now and again many erratics, large and small, and are, moreover, frequently disturbed and contorted as if from the grounding of pack-ice.

The phenomena which I have thus briefly sketched suffice to show that the epoch of local glaciation is to be clearly distinguished from that of the latest general _mer de glace_. I have long suspected, indeed, that the two may have been separated by as wide an interval of time as that which divided the earlier from the later epoch of general glaciation. Again and again I have searched underneath the terminal moraines, in the faint hope of detecting interglacial accumulations. My failure to discover these, however, did not weaken my conviction, for it was only by the merest chance that interglacial beds could ever have been preserved in such places. I feel sure, however, that they must occur among the older alluvia of our Lowlands. Indeed, as I shall point out in the sequel, it is highly probable that they are already known, and that we have hitherto failed to recognise their true position in the glacial series.

Although we have no direct evidence to prove that a long interglacial epoch of mild conditions immediately preceded the advent of our local ice-sheets and large valley-glaciers, yet the indirect evidence is so strong that we seem driven to admit that such must have been the case. To show this I must briefly recapitulate what is now known as to the glacial succession on the Continent. It has been ascertained, then, that the Scandinavian ice has invaded the low-grounds of Germany on two separate occasions, which are spoken of by Continental geologists as the "first" and "second" glacial epochs. The earlier of these was the epoch of maximum glaciation, when the inland ice flowed south into Saxony, and overspread a vast area between the borders of the North Sea and the base of the Ural Mountains. This ice-sheet unquestionably coalesced with the _mer de glace_ of the British Islands. Its bottom-moraine and the associated fluvio-glacial detritus are known in Germany as "Lower Diluvium," and the various phenomena connected with it clearly show that the inland-ice radiated outwards from the high-grounds of Scandinavia. The terminal front of that vast _mer de glace_ is roughly indicated by a line drawn from the south coast of Belgium round the north base of the Harz, and by Leipzig and Dresden to Krakow, thence north-east to Nijnii Novgorod, and further north to the head-waters of the Dvina and the shores of the Arctic Sea near the Tcheskaia Gulf.

The lower diluvium is covered in certain places by interglacial deposits and an overlying upper diluvium--a succession clearly indicative of climatic changes. In the interglacial beds occur remains of _Elephas antiquus_ and other Pleistocene mammals, and a flora which denotes a genial temperate climate. One of the latest discoveries of interglacial remains is that of two peat-beds lying between the lower and upper diluvium near Grünenthal in Holstein.[CA] Among the abundant plant-relics are pines and firs (no longer indigenous to Schleswig-Holstein), aspen, willow, white birch, hazel, hornbeam, oak, and juniper. Associated with these are _Ilex_ and _Trapa natans_, the presence of which, as Dr. Weber remarks, betokens a climate like that of western middle Germany. Amongst the plants is a water-lily, which occurs also in the interglacial beds of Switzerland, but is not now found in Europe. The evidence furnished by this and other interglacial deposits in north Germany shows that, after the ice-sheet of the lower diluvium had melted away, the climate became as temperate as that now experienced in Europe. Another recent find of the same kind[CB] is the "diluvial" peat, etc., of Klinge, in Brandenburg, described by Professor Nehring. These beds have yielded remains of elk (_Cervus alces_), rhinoceros (species not determined), a small fox (?), and Megaceros. This latter is not the typical great Irish deer, but a variety (_C. megaceros_, var. _Ruffii_, Nehring). The plant-remains include pine, fir (_Picea excelsa_), hornbeam, warty birch (_Betula verrucosa_), various willows (_Salix repens_, _S. aurita_, _S. caprea_ [?], _S. cinerea_), hazel, poplar (?), common holly, etc. It is worthy of note that here also the interglacial water-lily (_Cratopleura_) of Schleswig-Holstein and Switzerland makes its appearance. Dr. Weber writes me that the facies of this flora implies a well-marked temperate insular climate (Seeklima). The occurrence of holly in the heart of the Continent, where it no longer grows wild, is particularly noteworthy. The evidence furnished by such a flora leads one to conclude that at the climax of the genial interglacial epoch, the Scandinavian snow-fields and glaciers were not more extensive than they are at present.

[CA] _Neues Jahrbuch f. Min. Geol. u. Palæont._, 1891, ii., pp. 62, 228; _Ibid._, 1892, i., p. 114.

[CB] _Naturwissenschaftliche Wochenschrift_, Bd. vii. (1892), No. 4, p. 31. The plants were determined by Dr. Weber, Professor Wittmack, and Herr Warnstorf. [More recent investigations have considerably increased our knowledge of this flora. See _Naturwissenschaftliche Wochenschrift_, Bd. vii. (1892), Nr. 24, 25. _Ausland_, 1892, Nr. 20; _Neues Jahrb. f. Min., etc._, 1893, Bd. i., p. 95.]

The presence of the upper diluvium, however, proves that such genial conditions eventually passed away, and that an ice-sheet again invaded north Germany. But this later invasion was not on the same scale as that of the preceding one. The geographical distribution of the upper diluvium and the position of large terminal moraines put this quite beyond doubt. The boulder-clay in question spreads over the Baltic provinces of Germany, extending south as far as Berlin,[CC] and west into Schleswig-Holstein and Denmark. At the climax of this later cold epoch glaciers occupied all the fiords of Norway, but did not advance beyond the general coast-line. Norway at that time must have greatly resembled Greenland--the inland-ice covering the interior of the country, and sending seawards large glaciers that calved their icebergs at the mouths of the great fiords. In the extreme south, however, the glaciers did not quite reach the sea, but piled up large terminal moraines on the coast-lands, which may be followed thence into Sweden in an easterly direction by the lower end of Lake Wener and through Lake Wetter. A similar belt of moraines marks out the southern termination of the ice-sheet in Finland. Between Sweden and Finland lies the basin of the Baltic, which, at the epoch in question was filled with ice, forming a great Baltic glacier. This glacier overflowed the Öland Islands, Gottland, and Öland, fanning out as it passed towards the south-west and west, so as to invade on the south the Baltic provinces of Germany, while in the north it traversed the southern part of Scania, and overwhelmed the Danish islands as it spread into Jutland and Schleswig-Holstein. The course of this second ice-sheet is indicated by the direction of transport of erratics, etc., and by the trend of rock-striæ and _roches moutonnées_, as well as by the position of its terminal and lateral moraines.

[CC] Not quite so far south. There is no reason to believe that the ice-sheet of the so-called Great Baltic Glacier advanced beyond the Baltic ridge. The upper boulder-clay south of that ridge is the ground-moraine of an earlier glaciation--the equivalent of our upper boulder-clay. See note, page 324. Nov. 1, 1892.

Such, then, is the glacial succession which has been established by geologists in Scandinavia, north Germany, and Finland. The occurrence of two glacial epochs, separated by a long interval of temperate conditions, has been proved. The evidence, however, does not show that there may not have been more than two glacial epochs. There are certain phenomena, indeed, connected with the glacial accumulations of the regions in question which strongly suggest that the succession of changes was more complex than is generally understood. Several years ago Dr. A. G. Nathorst adduced evidence to show that a great Baltic glacier, similar to that underneath which the upper diluvium was amassed, existed before the advent of the vast _mer de glace_ of the so-called "first glacial epoch,"[CD] and his observations have been confirmed and extended by H. Lundbohm.[CE] The facts set forth by them prove beyond doubt that this early Baltic glacier smoothed and glaciated the rocks in southern Sweden in a direction from south-east to north-west, and accumulated a bottom-moraine whose included erratics are equally cogent evidence as to the trend of glaciation. That old moraine is overlaid by the lower diluvium--_i.e._, the boulder-clay, etc., of the succeeding vast _mer de glace_ that flowed south to the foot of the Harz--the transport of the stones in the superjacent clay indicating a movement from NNE. to SSW., or nearly at right angles to the trend of the earlier Baltic glacier. It is difficult to avoid the conclusion that we have here to do with the products of two distinct ice-epochs. But hitherto no interglacial deposits have been detected between the boulder-clays in question. It might, therefore, be held that the early Baltic glacier was separated by no long interval of time from the succeeding great _mer de glace_, but may have been merely a stage in the development of the latter. It is at all events conceivable that before the great _mer de glace_ attained its maximum extension, it might have existed for a time as a large Baltic glacier. I would point out, however, that if no interglacial beds had been recognised between the lower and the upper diluvium, geologists would probably have considered that the last great Baltic glacier was simply the attenuated successor of the preceding continental _mer de glace_. But we know this was not so; the two were actually separated by a long epoch of genial temperate conditions.

[CD] "Beskrifning. till geol. Kartbl. Trolleholm": _Sveriges Geologiska Undersökning_, Ser. Aa., Nr. 87.

[CE] "Om de äldre baltiska isströmmen i södra Sverige": _Geolog. Förening. i Stockholm Förhandl._, Bd. x., p. 157.

There are certain other facts that may lead us to doubt whether in the glacial phenomena of the Baltic coast-lands we have not the evidence of more than two glacial epochs. Three, and even four, boulder-clays have been observed in east and west Prussia. They are separated, the one from the other, by extensive aqueous deposits, which are sometimes fossiliferous. Moreover, the boulder-clays in question have been followed continuously over considerable areas. It is quite possible, of course, that all those boulder-clays may be the product of one epoch, laid down during more or less considerable oscillations of an ice-sheet. In this view of the case the intercalated aqueous deposits would indicate temporary retreats, while the boulder-clays would represent successive re-advances of one and the same _mer de glace_. On the other hand, it is equally possible, if not more probable, that the boulder-clays and intercalated beds are evidence of so many separate glacial and interglacial epochs. We cannot yet say which is the true explanation of the facts. But these being as they are, we may doubt if German glacialists are justified in so confidently maintaining that their lower and upper diluvial accumulations are the products of the "first" and "second" glacial epochs. Indeed, as I shall show presently, the upper diluvium of north Germany and Finland cannot represent the second glacial epoch of other parts of Europe.

For a long time it has been supposed that the glacial deposits of the central regions of Russia were accumulated during the advance and retreat of one and the same ice-sheet. In 1888, however, Professor Pavlow brought forward evidence to show that the province of Nijnii Novgorod had been twice invaded by a general _mer de glace_. During the first epoch of glaciation the ice-sheet overflowed the whole province, while only the northern half of the same region was covered by the _mer de glace_ of the second invasion. Again, Professor Armachevsky has pointed out that in the province of Tchernigow two types of glacial deposits appear, so unlike in character and so differently distributed that they can hardly be the products of one and the same ice-sheet. But until recently no interglacial deposits had been detected, and the observations just referred to failed, therefore, to make much impression. The missing link in the material evidence has now happily been supplied by M. Krischtafowitsch.[CF] At Troïzkoje, in the neighbourhood of Moscow, occur certain lacustrine formations which have been long known to Russian geologists. These have been variously assigned to Tertiary, lower glacial, post-glacial, and pre-glacial horizons. They are now proved, however, to be of interglacial age, for they rest upon and are covered by glacial accumulations. Amongst their organic remains are oak (_Quercus pedunculata_), alder (_Alnus glutinosa_, _A. incana_), white birch, hazel, Norway maple (_Acer platanoides_), Scots fir, willow, water-lilies (_Nuphar_, _Nymphæa_), mammoth, pike, perch, _Anadonta_, wing-cases of beetles, etc. The character of the plants shows that the climate of central Russia was milder and more humid than it is to-day.

[CF] _Bull. de la Soc. Impér. des Naturalistes de Moskau_, No. 4, 1890.

It is obvious that the upper and lower glacial deposits of central Russia cannot be the equivalents of the upper and lower diluvium of the Baltic coast-lands. The upper diluvium of those regions is the bottom-moraine of the so-called great Baltic glacier. At the time that glacier invaded north Germany, Finland was likewise covered with an ice-sheet, which flowed towards the south-east, but did not advance quite so far as the northern shores of Lake Ladoga. A double line of terminal moraines, traced from Hango Head on the Gulf of Finland, north-east to beyond Joensuu, puts this beyond doubt.[CG] The morainic deposits that overlie the interglacial beds of central Russia cannot, therefore, belong to the epoch of the great Baltic glacier. They are necessarily older. In short, it is obvious that the upper and lower glacial accumulations near Moscow must be on the horizon of the lower diluvium of north Germany. And if this be so, then it is clear that the latter cannot be entirely the product of one and the same _mer de glace_. When the several boulder-clays described by Schröder and others as occurring in the Baltic provinces of Germany are reinvestigated, they may prove to be the bottom-moraines of as many distinct and separate glacial epochs.

[CG] Sederholm, _Fennia_, i., No. 7; Frosterus, _ibid._, iii., No. 8; Ramsay, _ibid._, iv., No. 2.

It may be contended that the glacial and interglacial deposits of central Russia are perhaps only local developments--that their evidence may be accounted for by the oscillations of one single _mer de glace_. This explanation, as already pointed out, has been applied to the boulder-clays and intercalated aqueous beds of the lower diluvium of north Germany, and the prevalent character of the associated organic remains makes it appear plausible. It is quite inapplicable, however, to the similar accumulations in central Russia. During the formation of the freshwater beds of Troïzkoje, no part of Russia could have been occupied by an ice-sheet; the climate was more genial and less "continental" than the present. Yet that mild interglacial epoch was preceded and succeeded by extremely arctic conditions. It is impossible that such excessive changes could have been confined to central Russia. Germany, and indeed all northern and north-western Europe, must have participated in the climatic revolutions.

So far, then, as the evidence has been considered, we may conclude that three glacial and two interglacial epochs at least have been established for northern Europe. If this be the case, then a similar succession ought to occur in our own islands; and a little consideration of the evidence already adduced will suffice to show that it does. It will be remembered that the lower and upper boulder-clays of the British Islands are the bottom-moraines of two separate and distinct ice-sheets, each of which in its time coalesced on the floor of the North Sea with the inland-ice of Scandinavia. It is obvious, therefore, that our upper boulder-clay cannot be the equivalent of the upper diluvium of the Baltic coast-lands, of Sweden, Denmark, and Schleswig-Holstein. De Geer and others have shown that while the great Baltic glacier was accumulating the upper diluvium of North Germany, etc., the inland-ice of Norway calved its icebergs at the mouths of the great fiords. Thus, during the so-called "second" glacial epoch of Scandinavian and German geologists, the Norwegian inland-ice did not coalesce with any British _mer de glace_. The true equivalent in this country of the upper diluvium is not our upper boulder-clay, but the great valley-moraines of our mountain-regions. It is our epoch of large valley-glaciers which corresponds to that of the great Baltic ice-flow. Our upper and lower boulder-clays are on the horizon of the lower diluvium of Germany and the glacial deposits of central Russia.

It will now be seen that the evidence in Britain is fully borne out by what is known of the glacial succession in the corresponding latitudes of the Continent. I had inferred that our epoch of large valley-glaciers formed a distinct stage by itself, and was probably separated from that of the preceding ice-sheet by a prolonged interval of interglacial conditions. One link in the chain of evidence, however, was wanting: I could not point to the occurrence of interglacial deposits underneath the great valley-moraines. But these, as we have seen, form a well-marked horizon on the Continent, and we cannot doubt that a similar interglacial stage obtained in these islands. We may feel confident, in fact, that genial climatic conditions supervened on the dissolution of the last great _mer de glace_ in Britain, and that the subsequent development of extensive snow-fields and glaciers in our mountain-regions was contemporaneous with the appearance of the last great Baltic glacier.

We need not be surprised that interglacial beds should be well developed underneath the bottom-moraine of that great glacier, while they have not yet been recognised below the corresponding morainic accumulations of our Highlands and Uplands. The conditions in the low-grounds of the Baltic coast-lands favoured their preservation, for the ice in those regions formed a broad _mer de glace_, under the peripheral areas of which sub-glacial erosion was necessarily at a minimum and the accumulation at a maximum. In our Scottish mountain-valleys, however, the very opposite was the case. The conditions obtaining there were not at all comparable to those that characterised the low-grounds of northern Germany, etc., but were quite analogous to those of Norway, where, as in our own mountain-regions, interglacial beds are similarly wanting. It is quite possible, however, that patches of such deposits may yet be met with underneath our younger moraines, and they ought certainly to be looked for. But whether they occur or not in our mountain-valleys, it is certain that some of the older alluvia of our Lowlands must belong to this horizon. Hitherto all alluvial beds that overlie our upper boulder-clay have been classified as post-glacial; but since we have ascertained that our latest _mer de glace_ was succeeded by genial interglacial conditions, we may be sure that records of that temperate epoch will yet be recognised in such Lowland tracts as were never reached by the glaciers of the succeeding cold epoch. Hence, I believe that some of our so-called "post-glacial" alluvia will eventually be assigned to an interglacial horizon. Amongst these may be cited the old peat and freshwater beds that rest upon the upper boulder-clay at Hailes Quarry, near Edinburgh. To the same horizon, in all probability, belong the clays, with Megaceros, etc., which occur so frequently underneath the peat-bogs of Ireland. An interesting account of these was given some years ago by Mr. Williams,[CH] who, as a collector of Megaceros remains, had the best opportunity of ascertaining the nature of the deposits in which these occur. He gives a section of Ballybetagh Bog, nine miles south-east of Dublin, which is as follows:--

1. Boulder-clay.
2. Fine tenacious clay, without stones.
3. Yellowish clay, largely composed of vegetable matter.
4. Brownish clay, with remains of Megaceros.
5. Greyish clay.
6. Peat.

[CH] _Geol. Mag._, 1881, p. 354.